Personal navigation systems capable of providing highly accurate location information in global positioning system (GPS) denied environments or GPS-disrupted environments are sought after for military, first responder, and consumer applications. These personal navigation systems need to provide accurate position information when GPS is unavailable or unreliable for long periods of time (e.g., hours to days). GPS interruption can occur due to GPS line-of-sight blockage (e.g., buildings, forest canopy, caves, etc.) or due to electrical interference/jamming.
Typically, personal navigation systems use an inertial measurement unit (IMU), or some subset of inertial sensors, to measure changes in position and heading to track the movement of a person, ground vehicle, or air vehicle. Since IMU errors accumulate rapidly, additional sensors such as a compass, pressure sensor, or velocity sensors are added to constrain error growth and drift. Furthermore, algorithms based on motion classification or zero velocity update (ZUPT) are used to compensate and constrain distance error growth, but do not adequately constrain heading error. In order to bound heading errors a compass is often used with the ZUPT, however, compass accuracy still results in limited position performance and is inadequate for long, precise GPS-denied missions. Vision-based systems, using either optical flow or image/landmark recognition, can compensate for heading error, but tend to be computationally demanding.
Personal dead reckoning systems for navigating in GPS-denied environments have also been developed. Such systems, which are based on a fusion of inertial sensors, a compass, and a pressure sensor, are limited in accuracy to about 1-5% error over distance traveled. Distance error typically accounts for about 30% of total position error and heading error accounts for about 70% of the total position error.